[Radiance-general] RE: 60Watt bulb comparison

[John Mardaljevic]

Anthony,
 
> I had to construct a small black box (0.5 cube) for calibrating electronic
> sensors I am currently using to monitor a full scale light pipe
> installation, which in turn will be compared with a Radiance model thereof.

Yep, 0.5 (m) is small compared to the bulb diameter.
 
> Furthermore, Recent research (Maamari et al. 2003) shows that a cube dark
> box with an
> opening central to one of the walls is most appropriate for simulation
> modelling
> comparison because reflectance values are minimal and easily measured, and
> modelling complexities and associated anomalies are minimised.

Actually, I believe this scenario was used becuase it was simple enough
for there to be an exact analytical solution.  It's not really "appropriate"
at all for any real world example.
 
> Anyway, I am more interested in the accuracy of the program with respect to
> *daylight* rather than electric lighting. To date, I can only locate your
> own work to the accuracy of the Radiance accuracy with daylight (as in the
> RwR manual)...

What can I say?  The validation I carried out is reckoned to be the most
rigorous yet for daylight modelling.  Enough people have said so that I'm
inclined to believe them.  Of course, this is largely becuase I had access
to an excellent (still unique?) test dataset from the BRE.  It's described
here in gory detail:

http://www.iesd.dmu.ac.uk/~jm/zxcv-thesis/

There's no question that Radiance is *capable* of high accuracy predictions
provided that it is driven correctly - at least for the great number of plausible
architectural scenarios that are not too dissimilar from the validation scenario.

The tricky part is determining the extent of what I call the "domain of validity"
of the validation scenario.  In other words, how different from the validation scenario
does a scene have to be so that we begin to question the software's capacity to
deliver an accurate result?  (Rather than our own ability to describe the scene
with sufficient precision or drive the simulation correctly.) This and other
validation topics are addressed in the paper:

"Verification of Program Accuracy for Illuminance Modelling: Assumptions, Methodology
and an Examination of Conflicting Findings (2004) Lighting Res. Technol. (in press)

It should be out in a few months.

-John

-----------------------------------------------
Dr. John Mardaljevic                     
Senior Research Fellow
Institute of Energy and Sustainable Development
De Montfort University
The Gateway
Leicester
LE1 9BH, UK
+44 (0) 116 257 7972   
+44 (0) 116 257 7981 (fax)

jm at dmu.ac.uk   
http://www.iesd.dmu.ac.uk/~jm